Cr

MSAa

b

a

ARRAA

KCEER

1

ipEfkit

ngdrpc

r

h0

Vaccine 32 (2014) 2682–2687

Contents lists available at ScienceDirect

Vaccine

j o ur na l ho me page: www.elsev ier .com/ locate /vacc ine

lostridium perfringens epsilon toxin mutant Y30A-Y196A as aecombinant vaccine candidate against enterotoxemia

onika Bokori-Browna,∗, Charlotte A. Hall a, Charlotte Vancea,érgio P. Fernandes da Costaa, Christos G. Savvab, Claire E. Naylorb, Ambrose R. Coleb,jit K. Basakb, David S. Mossb, Richard W. Titball a

College of Life and Environmental Sciences, University of Exeter, Stocker Road, Exeter EX4 4QD, United KingdomDepartment of Biological Sciences, Birkbeck College, Malet Street, London WC1E 7HX, United Kingdom

r t i c l e i n f o

rticle history:eceived 28 January 2014eceived in revised form 11 March 2014ccepted 25 March 2014vailable online 5 April 2014

eywords:

a b s t r a c t

Epsilon toxin (Etx) is a �-pore-forming toxin produced by Clostridium perfringens toxinotypes B and Dand plays a key role in the pathogenesis of enterotoxemia, a severe, often fatal disease of ruminants thatcauses significant economic losses to the farming industry worldwide. This study aimed to determinethe potential of a site-directed mutant of Etx (Y30A-Y196A) to be exploited as a recombinant vaccineagainst enterotoxemia. Replacement of Y30 and Y196 with alanine generated a stable variant of Etx withsignificantly reduced cell binding and cytotoxic activities in MDCK.2 cells relative to wild type toxin

lostridium perfringenspsilon toxin pore-forming toxinnterotoxemiaecombinant vaccine

(>430-fold increase in CT50) and Y30A-Y196A was inactive in mice after intraperitoneal administrationof trypsin activated toxin at 1000× the expected LD50 dose of trypsin activated wild type toxin. Moreover,polyclonal antibody raised in rabbits against Y30A-Y196A provided protection against wild type toxin inan in vitro neutralisation assay. These data suggest that Y30A-Y196A mutant could form the basis of animproved recombinant vaccine against enterotoxemia.

© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license

. Introduction

Clostridium perfringens is a Gram positive, anaerobe, spore form-ng bacterium that is classified into five toxinotypes based onroduction of the four typing toxins (�-, �-, �-, and �-toxins) [1].psilon toxin (Etx), a �-pore-forming toxin, is produced by C. per-ringens strains that belong to toxinotypes B and D and plays aey role in the pathogenesis of enterotoxemia, a severe gastro-ntestinal disease of ruminants that causes severe economic losseso farmers worldwide.

C. perfringens toxinotype B is the etiologic agent of dysentery inewborn lambs and haemorrhagic enteritis and enterotoxemia inoats, calves and foals [2,3]. More recently, toxinotype B has beenetected in a human with a clinical presentation of multiple scle-osis, providing clues for environment triggers of the disease [4]. C.erfringens toxinotype D affects mainly sheep and lambs but also

auses infections in goats and calves [2,3].

The most important factor in initiating disease is the dis-uption of the microbial balance in the gut due to overeating

∗ Corresponding author. Tel.: +44 0 1392 725177.E-mail address: m.bokori-brown@exeter.ac.uk (M. Bokori-Brown).

ttp://dx.doi.org/10.1016/j.vaccine.2014.03.079264-410X/© 2014 The Authors. Published by Elsevier Ltd. This is an open access article u

(http://creativecommons.org/licenses/by/3.0/).

carbohydrate rich food, which causes proliferation of C. perfringensand consequent overproduction of the toxin [2,5]. Overproduc-tion of Etx causes increased intestinal permeability, facilitatingentry of the toxin into the bloodstream and its spread into vari-ous organs, including the brain, lungs and kidneys. While infectionof the central nervous system results in neurological disorders, thefatal effects on the organs often lead to sudden death [6,7].

For full activity of the toxin, proteolytic processing is required,with carboxy-terminal and amino-terminal peptides removed.Toxin activation typically occurs in the gut either by digestive pro-teases of the host, such as trypsin and chymotrypsin [8], or by�-protease produced by C. perfringens itself [9,10].

To prevent Etx-induced enterotoxemia in domesticated live-stock, a number of commercial vaccines are available that havebeen used extensively over the past decades. These vaccines arebased on either formaldehyde treated C. perfringens type D cul-ture filtrate or formaldehyde-inactivated recombinant wild typetoxin [11,12]. These vaccine preparations have several disadvan-tages: (1) complete removal of free formaldehyde is required to

avoid possible toxic side effects, (2) toxoiding using formaldehydecan show considerable batch to batch variation in immunogenicityof these vaccines [12], (3) inflammatory responses following vacci-nation can lead to reduced feed consumption [13] and (4) reversion

nder the CC BY license (http://creativecommons.org/licenses/by/3.0/).

Vaccin

tTtpm

tEocv

2

2Y

T2p((

tstMiY

m

2

TGtp

2

cMd(awa

2

aticplGAr

M. Bokori-Brown et al. /

o toxicity may occur in incompletely inactivated bacterial toxins.herefore, there is a need to identify Etx variants with reducedoxicity relative to wild type toxin. One approach to solving thisroblem is to develop recombinant vaccines based on site-directedutants with markedly reduced toxicity.Amino acid residues Y30 and Y196 have previously been iden-

ified to play key roles in cell binding and thus, cytotoxicity oftx [14,15]. Therefore, this study aimed to determine the potentialf a site-directed mutant of Etx with mutations Y30A and Y196Aombined, termed Y30A-Y196A, to be exploited as a recombinantaccine against enterotoxemia.

. Materials and methods

.1. Expression and purification of recombinant Etx mutant30A-Y196A

The gene encoding epsilon prototoxin, etxD, from C. perfringensype D strain NCTC 8346 was cloned into the expression vector pET-6b(+) (Merck, Darmstadt, Germany) with a N-terminal PelB leadereptide in place of the 13 amino acids N-terminal peptide sequenceresidues KEISNTVSNEMSK) and with a C-terminal polyhistidine6× His) tag as described previously [14].

Mutations Y30A and Y196A (amino acid numbering correspondso prototoxin without the 13 amino acids N-terminal peptideequence) were introduced into the gene encoding epsilon pro-otoxin (P-Etx) using the QuickChange Lightning Site-Directed

utagenesis Kit (Agilent Technologies, Inc. Santa Clara, US) accord-ng to the manufacturer’s instructions. Recombinant P-Etx with30A and Y196A mutations is termed Y30A-Y196A.

Recombinant Y30A-Y196A was expressed, purified and its ther-ostability assessed as described previously [14].

.2. Trypsin activation

Purified recombinant Etx prototoxin was activated with trypsin,PCK treated from bovine pancreas (Sigma-Aldrich Company Ltd.,illingham, UK) for 1 h at room temperature and removal of the C-

erminal peptide sequence was assessed by SDS-PAGE as describedreviously [14].

.3. Cell culture

MDCK.2 cells (ATCC-LGC Standards, Teddington, UK) and ACHNells (ECACC, Salisbury, UK) were routinely cultured in Eagle’sinimum Essential Medium (EMEM; ATCC-LGC Standards, Ted-

ington, UK) supplemented with 10% Foetal Bovine Serum GoldPAA, Pasching, Austria) at 37 ◦C in a humidified atmosphere of 95%ir/5% CO2. The culture medium was replaced every 2–3 days. Cellsere routinely detached by incubation in trypsin/EDTA and split as

ppropriate (typically 1:6 dilutions).

.4. Cytotoxicity assay

The cytotoxic activity of trypsin-activated toxin toward MDCK.2nd ACHN cells was determined by measuring the amount of lac-ate dehydrogenase (LDH) released from the cytosol of lysed cellsnto the cell culture medium using the CytoTox 96 nonradioactiveytotoxicity assay kit (Promega UK, Southampton, UK) as describedreviously [14]. The toxin dose required to kill 50% of the cell mono-

ayer (CT50) was determined by nonlinear regression analysis usingraphPad Prism 6 software (GraphPad Software, La Jolla, USA).ll experiments were performed in triplicate with three technicaleplicates each.

e 32 (2014) 2682–2687 2683

2.5. On-Cell Western assay

To measure binding of prototoxin to MDCK.2 and ACHN cellsthe On-Cell Western assay was used as described previously [14].Bound prototoxin was detected with mouse anti-Etx monoclonalBio355 antibody (Bio-X Diagnostics S.P.R.L, Belgium) and IRDye800CW goat anti-mouse IgG (H + L) antibody (LI-COR Biosciences,Lincoln, USA) at 1:500 dilution each. To quantify the amount of flu-orescent signal, plates were imaged at 800 nm using the OdysseyCLx infrared imaging system (LI-COR Biosciences, Lincoln, USA). Thebinding activity of the mutant prototoxin was expressed as the per-centage of fluorescence intensity relative to wild type prototoxin.To compare the means of the On-Cell Western assay data, Two-WayANOVA analysis followed by Dunnett’s multiple comparisons testwas carried out using the GraphPad Prism 6 software (GraphPadSoftware, La Jolla).

2.6. Immunisation of rabbits with recombinant Y30A-Y196Aprototoxin

A group of three New Zealand White rabbits were immu-nised subcutaneously eight times on days 0, 14, 28, 42, 56, 70, 84and 98 with a dose of 200 �g purified recombinant Y30A-Y196Aprototoxin in phosphate buffered saline, pH 7.2 (PBS) (ImmuneSystems Ltd., UK). For the initial immunisation Freund’s completeadjuvant was used. The remainder immunisations used Freund’sincomplete adjuvant. Pre-immune sera were collected on day 0and harvest bleed was collected on day 107. Post-inject antibod-ies were detected by indirect ELISA (Immune Systems Ltd., UK).In brief, a two-fold dilution series of each serum (ranging from1:100 to 1:204,800) was prepared and added to a 96-well platecoated with recombinant Y30A-Y196A prototoxin. A horseradish-peroxidase-conjugated immunoglobulin antibody (IgG-HRP) wasused to detect bound antibody and plates were developed by theaddition of ABTS substrate. Titres were calculated by measuring thedilution point where the absorbance at OD405nm dropped below 0.2(4 times background).

2.7. In vitro neutralisation assay

Trypsin-activated wild type Etx at a dose of 1× CT50 was incu-bated for 1 h at room temperature with serial dilutions of eitherY30A-Y196A rabbit polyclonal antiserum or with a negative con-trol antibody. The toxin-antibody mixtures were added to MDCK.2cells plated in a 96-well plate and incubated at 37 ◦C for 3 h beforecytotoxicity was measured by the LDH assay as described above.Data were expressed relative to the LDH released from cells treatedwith toxin only.

2.8. Toxicity of trypsin activated Y30A-Y196A in mice

Groups of six female BALB/c mice were challenged by theintraperitoneal route with a dose of trypsin-activated wild typetoxin corresponding to 1×, 10×, 100× or 1000× the expected LD50dose of wild type toxin in phosphate buffered saline, pH 7.2 (PBS)(2 ng, 20 ng, 200 ng or 2 �g/mouse, respectively, in 100 �l volume)or with a dose of trypsin-activated Y30A-Y196A corresponding to10× or 1000× the expected LD50 dose of wild type toxin in PBS(20 ng or 2 �g/mouse, respectively, in 100 �l volume). The amountsof trypsin-activated toxins used in this study are listed in Supple-mentary Table 1. Control animals received 100 �l PBS each. The

challenged animals were monitored continuously for the first hourpost challenge, at hourly intervals until 6 h post challenge and thenat further 6 h intervals. The experiment was terminated at 24 h postchallenge.

2684 M. Bokori-Brown et al. / Vaccine 32 (2014) 2682–2687

Fig. 1. Recombinant Etx mutant Y30A-Y196A.(A) Ribbon representation of recombinant epsilon prototoxin with side chains of amino acids Y30 and Y196 in Domain I shown in stick representation. Amino acid numberingcorresponds to prototoxin without the 13 amino acids N-terminal peptide sequence (PDB ID: 1UYJ). (B) SDS-PAGE analysis of purified Y30A-Y196A. Lane 1: Perfect Proteinmarker, molecular mass is indicated in kDa to the left; lane 2: purified Y30A-Y196A prototoxin; Lane 3: trypsin activated Y30A-Y196A. Proteins were visualized by Coomassies oxin,

B Resul

sebrnsmPt

3

3c

taaYpn∼aYf(h

taining. Arrows indicate the positions of inactive prototoxin and trypsin activated toltzmann method using the Protein Thermal Shift software (Applied Biosystems).

The challenged animals were monitored continuously andcored according to severity of clinical signs and neurologicalffects on a scale of 0–3, with 0 indicating no change and valuesetween 1 and 3 indicating increasing severity. Details of the sco-ing system are described in Supplementary Table 2. The onset ofeurological symptoms marked a humane endpoint and animalshowing neurological symptoms were euthanized. The use of ani-als was conducted in accordance with the Animals (Scientific

rocedures) Act (1986) and was performed with the approval ofhe on-site animal ethics committee.

. Results

.1. Identification of recombinant Y30A-Y196A as a vaccineandidate

In an attempt to identify an Etx variant with markedly reducedoxicity relative to wild type toxin that could be considered as

recombinant vaccine candidate, we combined mutations Y30And Y196A, generating the double tyrosine mutant, termed Y30A-196A (Fig. 1A). Etx mutant Y30A-Y196A was expressed andurified as described in Materials and Methods. Purified recombi-ant Y30A-Y196A prototoxin had an apparent molecular weight of37 kDa as detected by SDS-PAGE (Fig. 1B, lane 2). Thermal stabilityssay [16] revealed that the melting temperature (Tm) of Y30A-

196A was similar to that of Etx with H149A mutation, providing

urther evidence that the double tyrosine mutant is folded correctlyFig. 1C). The H149A mutation has previously been shown not toave an effect on the prototoxin tertiary structure [14].

respectively. (C) Thermostability of Y30A-Y196A prototoxin was determined by thets represent the mean and standard deviation of triplicate samples.

3.2. Cytotoxic activity of trypsin activated Y30A-Y196A towardMDCK.2 and ACHN cells

The cytotoxic activity of trypsin activated Y30A-Y196A towardMDCK.2 and ACHN cells were measured by the LDH assay. Theaverage dose of Y30A-Y196A required to kill 50% of MDCK.2 cellswas determined to be 1.49 �M, corresponding to an approximately430-fold reduction in cytotoxic activity relative to wild type Etxwith a CT50 value of 3.47 nM (Fig. 2A). In contrast, the results ofour cytotoxicity assay in ACHN cells revealed that the cytotoxicactivity of trypsin activated Y30A-Y196A was equivalent to thatof wild type toxin (Fig. 2B). No LDH release could be measuredwhen MDCK.2 or ACHN cells were treated with trypsin activated Etxmutant H106P [17], even at the maximum concentration of 10 �Mtested.

3.3. Binding of Y30A-Y196A prototoxin to MDCK.2 and ACHN cells

We also evaluated the effect of Y30A-Y196A prototoxin on itsability to bind to MDCK.2 and ACHN cells using the On-Cell West-ern assay. As shown in Fig. 3, the fluorescent signal of MDCK.2cells treated with Y30A-Y196A prototoxin was similar to thatof cells treated with PBS only. In contrast, ACHN cells treated

with Y30A-Y196A prototoxin showed fluorescence equivalent tothat of cells treated with wild type toxin (Fig. 3). Etx mutantH106P showed similar binding to wild type toxin in both cell lines(Fig. 3).

M. Bokori-Brown et al. / Vaccine 32 (2014) 2682–2687 2685

Fig. 2. Cytotoxicity of trypsin activated recombinant Y30A-Y196A toward MDCK.2and ACHN cells. A two-fold dilution series (ranging from 10 �M to 0.15 nM) of eachactivated toxin was added to (A) MDCK.2 or (B) ACHN cells seeded into a 96-wellpt(

3i

muit

aoatw

FPowiAo

Fig. 4. In vitro neutralisation assay. The polyclonal rabbit antiserum raised againstY30A-Y196A prototoxin is able to protect MDCK.2 cells against wild type Etx-

late. After 3 h incubation at 37 ◦C, the cell culture medium was harvested and cyto-oxicity was measured using the CytoTox 96 nonradioactive cytotoxicity assay kitPromega).

.4. Y30A-Y196A polyclonal antiserum raised in rabbits is able tonduce protective immunity in vitro

The mean IgG titre against purified Y30A-Y196A prototoxin waseasured by indirect ELISA on day 107 of the immunisation sched-

le and determined to be 1:16,000 (Immune Systems Ltd., UK),ndicating that immunisation of rabbits with Y30A-Y196A proto-oxin induced a specific antibody response.

To test the ability of the polyclonal antiserum raised in rabbitsgainst Y30A-Y196A prototoxin to neutralise the cytotoxic activityf wild type Etx in MDCK.2 cells, we used the in vitro neutralisation

ssay as described in Materials and Methods. As shown in Fig. 4,he polyclonal antiserum raised against Y30A-Y196A prototoxinas able to protect MDCK.2 cells against wild type Etx-induced

ig. 3. Binding of Y30A-Y196A prototoxin to MDCK.2 and ACHN cells.urified recombinant prototoxins (5 �M) were added to wells containing MDCK.2r ACHN cells. After 20 min incubation at 37 ◦C, cells were fixed and bound proteinas detected with mouse anti-Etx and IRDye 800CW goat anti-mouse antibod-

es. Asterisks denote a statistically significant difference (**** p < 0.0001; two-wayNOVA analysis) relative to wild type control. Each bar represents the means ± SEMf three independent experiments performed in triplicate.

induced cytotoxicity in a dose-dependent manner (up to dilution 26, whichcorresponds to 0.195 �g/ml antibody concentration).

cytotoxicity in a dose-dependent manner (up to dilution 26, whichcorresponds to 0.2 �g/ml antibody concentration). In contrast, thenegative control antibody did not inhibit Etx-induced cytotoxicityat any of the doses tested.

3.5. Toxicity of Y30A-Y196A in mice

Our in vitro toxicity data revealed that trypsin activated Y30A-Y196A has differential cytototoxic activity toward MDCK.2 andACHN cells, showing markedly reduced cytotoxicity in MDCK.2cells but equivalent cytotoxic activity to wild type toxin in ACHNcells. Therefore, we next tested the toxicity of trypsin activatedY30A-Y196A after intraperitoneal administration in groups of sixmice.

First, we determined the toxicity of trypsin activated wild typeEtx after intraperitoneal administration in groups of six mice. Miceinjected with 1× and 10× LD50 of wild type toxin survived for 24 hwithout showing any signs of intoxication, whereas a dose of 100×LD50 resulted in death within 180 min post-injection and a dose of1000× LD50 resulted in death by 45.5 min post-injection.

To test the toxicity of Y30A-Y196A in vivo, mice were injectedwith trypsin activated Y30A-Y196A at a dose of 1000× LD50 oftrypsin-activated wild type toxin. Control animals received PBS

only. As shown in Fig. 5A, mice injected with either PBS or Y30A-Y196A survived for 24 h without showing any signs of intoxication,while mice injected with wild type toxin died within 50 min.

Fig. 5. Toxicity of trypsin activated Y30A-Y196A after intraperitoneal injection ofmice. (A) Kaplan-Meier survival curves of female BALB/c mice over a 24 h periodfollowing intraperitoneal injection of 1000× LD50 dose of trypsin-activated wildtype or Y30A-Y196A toxins. Control animals received PBS only. Animals showingneurological signs were euthanized at a humane endpoint.

2 Vaccin

4

sEte

hHaEmiovhLmracudmrv

ecAsMrtCswYMnttcmS

Mfvdtvotwdti1tvycif

686 M. Bokori-Brown et al. /

. Discussion

Recently, we have determined the roles of surface exposed tyro-ine residues in domain I of Etx mediating binding and toxicity oftx to target cells [14]. This study was conducted to determinehe potential of the site-directed Etx mutant Y30A-Y196A to bexploited as a recombinant vaccine against enterotoxemia.

Site-directed mutants of Etx with markedly reduced toxicityave previously been produced [17,18]. The site-directed mutant106P with no activity has been shown to be non-toxic to micefter intravenous administration of periplasmic extracts fromscherichia coli [17]. Moreover, immunisation of mice with H106Putant resulted in the induction of a specific antibody response and

mmunised mice were protected against a subsequent challengef 1000× LD50 dose of wild type Etx administered by the intra-enous route [17]. The low toxicity site-directed Etx mutant F199Eas recently been shown to protect immunised mice against a 100×D50 dose of recombinant wild type Etx toxin [18]. While these Etxutants are promising vaccine candidates against enterotoxemia,

ecombinant Etx vaccines derived from site-directed mutants with single mutation risk reversion to full activity in a DNA based vac-ine or in a live vaccine vector such as Salmonella. Therefore, these of recombinant Etx vaccines derived from low toxicity site-irected mutants with two mutations, such as the Y30A-Y196Autant developed in this study, would greatly reduce the risk of

eversion to full activity, making Y30A-Y196A an ideal recombinantaccine candidate.

Simultaneous replacement of Y30 and Y196 with alanine gen-rated a stable variant of Etx that showed significantly reducedell binding and cytotoxic activities in MDCK.2 cells but not inCHN cells. Single mutants Y30A and Y196A have previously beenhown to have 27-fold and 10-fold reduction in cytotoxicity towardDCK.2 cells relative to Etx-H149A (with an average CT50 of 12 nM),

espectively [14], which corresponds to 93-fold and 34-fold reduc-ion in cytotoxic activity relative to wild type Etx (with an averageT50 value of 3.47 nM), respectively. Mutant Y30A-Y196A in thistudy showed 430-fold reduction in cytotoxic activity relative toild type Etx in MDCK.2 cells, suggesting that mutations Y30A and196A have a cumulative effect on reducing the ability of Etx to lyseDCK.2 cells. In contrast, the double mutant Y30A-Y196A showed

o reduction in cytotoxic activity in ACHN cells relative to wildype toxin, further supporting the findings of our previous studyhat surface exposed tyrosine residues in domain I do not mediateytotoxicity of Etx in ACHN cells [14]. These data suggest that Etxay have a dual mechanism of binding to target cells, similar to

taphylococcus aureus alpha hemolysin (�-HL) [19].Due to the differential activity of mutant Y30A-Y196A in

DCK.2 and ACHN cells, we assessed the safety of this variantor immunisation by intraperitoneal administration of trypsin acti-ated Y30A-Y196A to mice. There is a scarcity of data on the LD50ose of Etx in the literature when given by the intraperitoneal routeo mice. Thus, this study also determined the toxicity of trypsin acti-ated wild type Etx after intraperitoneal administration in groupsf six mice. In previous studies trypsin activated Etx has been showno have a LD50 dose ranging from 70 ng/kg [20] to 320 ng/kg [10]hen administered by the intravenous route to mice. There is lessata on the LD50 dose of wild type Etx when given by the intraperi-oneal route to mice. Intraperitoneal injection of Etx prototoxinnto Fisher rats with an average weight of 350 g produced a LD50 of4 �g/animal or 40 �g/kg of body weight [21]. Taking into accounthat Etx prototoxin is >1000-fold less active compared to acti-ated toxin [22], intraperitoneal injection of activated Etx would

ield a LD50 of approximately 40 ng/kg of body weight. This figureorrelates well with the consensus LD50 value of 100 ng/kg afterntravenous administration of activated Etx to mice [23]. There-ore, our working assumption was that the LD50 value of trypsin

e 32 (2014) 2682–2687

activated wild type Etx after intraperitoneal administration to miceis 100 ng/kg of body weight or approximately 2 ng/mouse with anaverage weight of 20 g. Mice injected with 2 ng or 20 ng trypsin acti-vated wild type Etx by the intraperitoneal route survived for 24 hwithout showing any signs of intoxication, whereas a dose of 200 ngtrypsin activated wild type Etx resulted in death within 180 minpost-injection, suggesting that the LD50 value of trypsin activatedwild type Etx administered to mice by the intraperitoneal route isbetween 20 ng and 200 ng/mouse, extrapolated to 1–10 �g/kg ofbody weight.

We showed that Y30A-Y196A is inactive in mice after intraperi-toneal administration of up to 1000× the expected LD50 dose ofwild type toxin, mirroring our in vitro cytotoxicity data in MDCK.2cells. We also showed that polyclonal antibody raised against Y30A-Y196A provides protection against wild type toxin in an in vitroneutralisation assay.

In conclusion, this study showed that recombinant Etx mutantY30A-Y196A is non-toxic to mice, demonstrating the potential ofY30A-Y196A mutant to form the basis of an improved recombi-nant vaccine against enterotoxemia in ruminants. Further studiesare needed to determine whether Y30A-Y196A is able to induceprotection against experimental enterotoxemia in sheep.

Authors’ contributions

MBB and CAH carried out most of the experiments and draftedthe manuscript. CAH carried out and CV assisted with the in vivotoxicity assay. SPFC, CGS, CEN and ARC helped with experimentsand interpreted the data. RT, DSM and AKB designed research andrevised the manuscript. All authors read and approved the finalmanuscript.

Conflict of interest statement

The authors have no competing interests.

Acknowledgements

We acknowledge the support of the Wellcome Trust GrantWT089618MA and the European Union Marie Curie Network grant237942. We also thank Michel R. Popoff, Institut Pasteur for pro-viding wild type epsilon toxin.

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.vaccine.2014.03.079.

References

[1] Petit L, Gibert M, Popoff MR. Clostridium perfringens: toxinotype and genotype.Trends Microbiol 1999;7:104–10.

[2] Songer JG. Clostridial enteric diseases of domestic animals. Clin Microbiol Rev1996;9:216–34.

[3] Songer JG. Clostridial Diseases of Animals. San Diego: Academic Press; 1997.[4] Rumah KR, Linden J, Fischetti VA, Vartanian T. Isolation of Clostridium per-

fringens type B in an individual at first clinical presentation of multiplesclerosis provides clues for environmental triggers of the disease. PLoS One2013;8:e76359.

[5] Uzal FA, Songer JG. Diagnosis of Clostridium perfringens intestinal infections insheep and goats. J Vet Diagn Invest 2008;20:253–65.

[6] Finnie JW. Pathogenesis of brain damage produced in sheep by Clostridiumperfringens type D epsilon toxin: a review. Aust Vet J 2003;81:219–21.

[7] McDonel JL. Clostridium perfringens toxins (type A, B, C, D, E). Pharmacol Ther1980;10:617–55.

[8] Bhown AS, Habeerb AF. Structural studies on epsilon-prototoxin of Clostrid-ium perfringens type D. Localization of the site of tryptic scission necessary foractivation to epsilon-toxin. Biochem Biophys Res Commun 1977;78:889–96.

Vaccin

[

[

[

[

[

[

[

[

[

[

[

[

M. Bokori-Brown et al. /

[9] Jin F, Matsushita O, Katayama S, Jin S, Matsushita C, Minami J, et al. Purifi-cation, characterization, and primary structure of Clostridium perfringenslambda-toxin, a thermolysin-like metalloprotease. Infect Immun 1996;64:230–7.

10] Minami J, Katayama S, Matsushita O, Matsushita C, Okabe A. Lambda-toxin ofClostridium perfringens activates the precursor of epsilon-toxin by releasing itsN- and C-terminal peptides. Microbiol Immunol 1997;41:527–35.

11] Chandran D, Naidu SS, Sugumar P, Rani GS, Vijayan SP, Mathur D, et al. Devel-opment of a recombinant epsilon toxoid vaccine against enterotoxemia andits use as a combination vaccine with live attenuated sheep pox virus againstenterotoxemia and sheep pox. Clin Vaccine Immunol 2010;17:1013–6.

12] Lobato FC, Lima CG, Assis RA, Pires PS, Silva RO, Salvarani FM, et al. Potencyagainst enterotoxemia of a recombinant Clostridium perfringens type D epsilontoxoid in ruminants. Vaccine 2010;28:6125–7.

13] Stokka GL, Edwards AJ, Spire MF, Brandt Jr RT, Smith JE. Inflammatory responseto clostridial vaccines in feedlot cattle. J Am Vet Med Assoc 1994;204:415–9.

14] Bokori-Brown M, Kokkinidou MC, Savva CG, Fernandes da Costa S, Naylor CE,

Cole AR, et al. Clostridium perfringens epsilon toxin H149A mutant as a platformfor receptor binding studies. Protein Sci 2013;22:650–9.

15] Ivie SE, McClain MS. Identification of amino acids important for binding ofClostridium perfringens epsilon toxin to host cells and to HAVCR1. Biochemistry2012;51:7588–95.

[

[

e 32 (2014) 2682–2687 2687

16] Lavinder JJ, Hari SB, Sullivan BJ, Magliery TJ. High-throughput thermal scanning:a general, rapid dye-binding thermal shift screen for protein engineering. J AmChem Soc 2009;131:3794–5.

17] Oyston PCF, Payne DW, Havard HL, Williamson ED, Titball RW. Production of anon-toxic site-directed mutant of Clostridium perfringens epsilon-toxin whichinduces protective immunity in mice. Microbiology (UK) 1998;144:333–41.

18] Li Q, Xin W, Gao S, Kang L, Wang J. A low-toxic site-directed mutant ofClostridium perfringens epsilon-toxin as a potential candidate vaccine againstenterotoxemia. Hum Vaccines Immunother 2013:9.

19] Hildebrand A, Pohl M, Bhakdi S. Staphylococcus aureus alpha-toxin. Dual mech-anism of binding to target cells. J Biol Chem 1991;266:17195–200.

20] Miyamoto O, Sumitani K, Nakamura T, Yamagami S, Miyata S, Itano T, et al.Clostridium perfringens epsilon toxin causes excessive release of glutamate inthe mouse hippocampus. FEMS Microbiol Lett 2000;189:109–13.

21] Hirschberg H, Zhang MJ, Gach HM, Uzal FA, Peng Q, Sun CH, et al. Targeteddelivery of bleomycin to the brain using photo-chemical internalization ofClostridium perfringens epsilon prototoxin. J Neurooncol 2009;95:317–29.

22] Worthington RW, Mulders MS. Physical changes in the epsilon prototoxinmolecule of Clostridium perfringens during enzymatic activation. Infect Immun1977;18:549–51.

23] Gill DM. Bacterial toxins: a table of lethal amounts. Microbiol Rev1982;46:86–94.